Download Determination of CYP2D6 Gene Copy Number

Clinical Chemistry 51:3
522–531 (2005)
Molecular Diagnostics
and Genetics
Determination of CYP2D6 Gene Copy Number
by Pyrosequencing
Erik Söderbäck,1† Anna-Lena Zackrisson,2 Bertil Lindblom,2 and Anders Alderborn1*
Background: Identification of CYP2D6 alleles *5 (deletion of the whole CYP2D6 gene) and *2xN (gene duplication) is very important because they are associated with
decreased or increased metabolism of many drugs. The
most commonly used method for analysis of these alleles
is, however, considered to be laborious and unreliable.
Methods: We developed a method to determine the
copy number of the CYP2D6*5 and CYP2D6*2xN alleles
by use of PyrosequencingTM technology. A single set
of PCR and sequencing primers was used to coamplify
and sequence a region in the CYP2D6 gene and
the equivalent region in the CYP2D8P pseudogene,
and relative quantification between these fragments
was performed. The CYP2D8P-specific Pyrosequencing
peak heights were used as references for the CYP2D6specific peak heights.
Results: Analysis of 200 pregenotyped samples showed
that this approach reliably resolved 0 – 4 genome copies
of the CYP2D6 gene. In 15 of these samples, the peak
pattern from one analyzed position was unexpected but
could be solved by conclusive results from a second
position. The method was verified on 270 other samples,
of which 267 gave results that corresponded to the
expected genotype. One of the samples could not be
interpreted. The reproducibility of the method was high.
Conclusions: CYP2D6 gene copy determination by Pyrosequencing is a reliable and rapid alternative to other
methods. The use of an internal CYP2D8P control as
well as generation of a sequence context ensures a
robust method and hence facilitates method validation.
Cytochrome P450 (CYP) is a family of proteins involved
in the degradation of many toxic compounds. The human
genome contains more than 60 unique CYP genes, of
which 6 are considered the most metabolically active (1 ).
Many drugs, being chemically directly or closely related
to naturally toxic compounds, are targets for CYP enzymatic activities. It is well known that polymorphisms in
the CYP genes affect the activities and/or specificities of
the encoded enzymes, which causes differences in the
response to drugs degraded by these enzymes.
One of the most important CYP enzymes with respect
to drug degradation is encoded by the CYP2D6 gene
[GenBank accession no. M33388 (2 )], which has been
shown to be involved in the degradation of ⬃100 drug
compounds (3 ). Approximately 80 different alleles of the
CYP2D6 gene have been identified (4 ), and their presence
varies between ethnic populations. The enzymatic activities in persons carrying these allelic differences vary from
total absence to different degrees of ultrarapid metabolism, which causes considerable variability in the response to certain drug treatments.
The CYP2D6 gene is located on chromosome 22, together with the two pseudogenes CYP2D7P [GenBank
accession nos. X58467 and X58468 (5 )] and CYP2D8P
[GenBank accession no. M33387 (2 )], which are localized
in tandem. Low or eliminated CYP2D6 enzyme activity is
usually caused by point mutations, small deletions, or
insertions. The CYP2D6*5 allele, however, carries a deletion of the complete gene. This allele is one of the most
frequent CYP2D6 alleles, and homozygous carriers totally
lack metabolic activity. On the other hand, rapid metabolizers often carry a duplication, or even multiplication (up
to 12 copies have been found), of the CYP2D6 gene, most
commonly the CYP2D6*2 allele [CYP2D6*2xN (1, 6, 7 )].
The increased enzymatic activity in individuals carrying
duplications enhances the degradation of drugs, which
may cause nontherapeutic drug concentrations at typical
dosages. Many antidepressants, for example, are metabolized by CYP2D6, and there is a distinct relationship
between the number of CYP2D6 genes and the rate of
metabolism of antidepressants (8 ). The presence of the
CYP2D6 duplication shows an interesting north–south
© 2005 American Association for Clinical Chemistry
1
Biotage AB, Uppsala, Sweden.
National Board of Forensic Medicine, Department of Forensic Genetics,
Linköping, Sweden.
†Current address: Biology Education Centre, Uppsala University, Box 592,
SE-751 24 Uppsala, Sweden.
*Address correspondence to this author at: Biotage AB, Kungsgatan 76,
SE-753 18 Uppsala, Sweden. Fax 46-18-591922; e-mail anders.alderborn@
telia.com.
Received September 20, 2004; accepted December 27, 2004.
Previously published online at DOI: 10.1373/clinchem.2004.043182
2
522
Clinical Chemistry 51, No. 3, 2005
gradient in the European population, with a frequency of
⬃1% in Scandinavia (9 ) to 7–10% in Italy (10 ) and as high
as 29% in Ethiopia in northern Africa (11 ).
Several methods for determination of CYP2D6 gene
copy number have been reported. Although considered as
relatively laborious and unreliable, the most commonly
used method is a long-range PCR spanning the CYP-REP
regions flanking the CYP2D6 gene, followed by gel-based
size analysis of the PCR products (12 ). Other methods are
based on quantitative amplification of a small PCR fragment from the CYP2D6 gene, in which the amount of PCR
product reflects the number of CYP2D6 genes in the
analyzed genome (13, 14 ). The amount of CYP2D6-specific PCR product is assessed in relation to a coamplified
region from an unrelated housekeeping gene, which is
assumed to consistently be a single-copy gene.
PyrosequencingTM is a nonelectrophoretic, real-time
DNA sequencing technology (15, 16 ). A primer is hybridized to a single-stranded PCR template, and the sequencing analysis is started by addition of nucleotides. The
nucleotides are added sequentially, and through coupled
enzymatic reactions, the polymerase-catalyzed incorporation of nucleotides can be monitored as light peaks in a
PyrogramTM. Pyrosequencing has been used for many
applications, such as genotyping of single-nucleotide
polymorphisms (17, 18 ), mutation detection (19, 20 ), and
sequence detection (21, 22 ). The peak heights in the
Pyrogram are directly proportional to the emitted light
from the Pyrosequencing reaction. This nature of the light
peaks enables quantitative analyses such as assessment of
allele frequency in a mixed sample population (23–25 )
and degree of DNA methylation (26, 27 ).
Several assay protocols have been developed for analysis of CYP alleles by Pyrosequencing (28 –32 ). In this
report, we describe an easy single-well assay for determination of CYP2D6 gene copy number, based on relative
quantifications from Pyrosequencing reactions using the
pseudogene CYP2D8P as reference.
Materials and Methods
dna samples for assay development and
verification
A total of 200 DNA samples from different laboratories
were used as a key for development of an assay based on
Pyrosequencing technology. These DNA samples were
extracted by different methods and stored frozen until
use.
An additional set of DNA samples from 270 Swedish
blood donors were analyzed to verify the assay. The
Regional Ethics Committee had given permission for
the analysis, and all donors had given their consent.
These DNA samples were extracted by the dodecyltrimethylammonium bromide/cetyltrimethylammonium bromide (DTAB/CTAB) method (33 ) and pregenotyped for
the CYP2D6*2, *3, *4, and *6 alleles by a specific Pyrosequencing method (29 ). Individuals hypothetically ho-
523
mozygous for CYP2D6*1, *2, *3, *4, or *6 were tested for
CYP2D6*5 by a multiplex long PCR (7 ) with modifications (29 ). Analysis for gene duplication was performed
in 26 of the 270 samples with an established long-fragment PCR method (34 ) with several modifications: Primers were manufactured (Invitrogen AB) according to
published primer sequences (34 ). A 25-␮L PCR reaction
was carried out with the Expand Long Template PCR
System (Roche Molecular Biochemicals). The reaction
mixture contained 2.5 ␮L of 10⫻ PCR buffer 3 (22.5 mM
MgCl2), 5 ␮L of deoxynucleotide triphosphate mixture
(2.5 mM each nucleotide), 0.36 ␮L of enzyme mixture
(3.5 U/␮L), 0.4 ␮L of each primer (cyp-17F, cyp-32R, and
cyp-207 F; 20 ␮M), 11.3 ␮L of water, and 5 ␮L of genomic
DNA (10 ng/␮L). The conditions for amplification with
the primer pairs cyp-17F/cyp-32R and cyp-207F/cyp-32R
were as follows: an initial denaturing step of 94 °C for 2
min, followed by 35 cycles of 94 °C for 15 s, 64 °C for 30 s,
and 68 °C for 6 min, with a final elongation step of 68 °C
for 7 min. The PCR products were separated and detected
in ethidium bromide-containing 0.8% agarose gels. The
number of tandem copies of the CYP2D6 gene could not
be determined by the long-PCR method used.
assay design
The three genes CYP2D6, CYP2D7P, and CYP2D8P were
aligned to find suitable regions for PCR primer positioning
(to amplify CYP2D6 and CYP2D8P and to avoid amplification of CYP2D7P). PCR designs in which only the CYP2D6
and the CYP2D8P genes were amplified were further analyzed for optimization of Pyrosequencing specificity. One or
several sequencing primers for each PCR fragment were
located upstream of regions with nucleotide differences
between the CYP2D6 and CYP2D8P genes. In the Pyrosequencing analysis, gene-specific signals were achieved by
design of an assay-specific nucleotide addition scheme. The
CYP2D6*5/CYP2D6*2xN assay was thus based on relative
quantification of Pyrosequencing peak signals between a
CYP2D6 gene-specific amplicon and a CYP2D8P amplicon.
The Pyrosequencing reactions were designed so that possible unwanted amplification of the highly similar CYP2D7P
gene did not interfere with any of the CYP2D6/CYP2D8Pspecific peaks.
primer sequences
Primers from three different vendors were successfully
tested in the assay. The upstream PCR primer used was
A1058FP (5⬘-GGTGGCTGACCTGTTCTYTG-3⬘). The sequence includes three internal sites where the common
D6/D8 sequence diverges from the D7 sequence. Keeping
the cycling annealing temperature above 58 °C excludes
amplification of the CYP2D7P gene. The degeneration at
the third position from the 3⬘ end of the primer (bolded Y)
is attributable to the variation between the CYP2D6 and
CYP2D8P sequences (C/T). The downstream PCR primer
was biotinylated A1051RPB (5⬘-GGGCTCACGCTGCACATC-3⬘). One of the sequencing primers used was
524
Söderbäck et al.: CYP2D6 Gene Copy Number
Fig. 1. Alignment of the CYP2D6 gene at the end of exon 6 with the CYP2D7P and CYP2D8P pseudogenes (GenBank accession nos. M33388 for
CYP2D6, X58467 for CYP2D7P, and M33387 for CYP2D8P).
Arrows indicate PCR primers (A1058FP and A1051RPB) and sequencing primers (A685FP and A1050FP). The shaded areas cover the bases generating peaks in the
Pyrosequencing reactions.
A685FP (5⬘-CGGGATGGTGACCACCTC-3⬘). The C at the
3⬘ end of the primer is homologous in the D6 and the D8
sequences but differs from the D7 sequence. The other
sequencing primer used was A1050FP (5⬘-GGCCTCCTGCTCATGATCC-3⬘).
pcr conditions
The PCR reagents used were all from the Qiagen HotStarTaq reagent set (cat. no. 203205; Qiagen) except for the
deoxynucleotide triphosphates (100 mM solution; cat. no.
27-2035-02; Amersham Biosciences), and autoclaved milliQ water was used for diluting the reagents. In the final
PCR mixture, the concentrations were 1.5 mM MgCl2,
2.5 mM each nucleotide, 200 nM each primer, and
0.2 ng/␮L template DNA. Addition of Q solution, added
according to the Qiagen manual, was necessary for PCR
product specificity. The PCR cycling was initiated with
15 min of enzyme activation at 95 °C (according to the
Qiagen instructions) followed by 45 cycles at 95 °C for
30 s, 59 °C for 30 s, and 72 °C for 30 s. The reaction was
finished with a 10-min final extension at 72 °C. The
quality of the PCR products was checked on agarose gels.
PCR instruments from Hybaid (Thermo-Hybaid Model
MBS 0.26) and Applied Biosystems (Models 6700 and
9600) generated PCR products successfully used in the
subsequent Pyrosequencing analysis. To test the influence
of PCR annealing temperature on Pyrosequencing peak
heights, measured as relative light units, we performed a
temperature gradient PCR on a Mastercycler gradient
(Eppendorf) with annealing temperatures ranging from
50 to 65 °C over the entire PCR block.
analysis
The PCR products were prepared for Pyrosequencing
analysis by use of a Vacuum Prep Workstation (Biotage
AB), and the sequencing reactions were performed on
either a PSQ 96MA System or a PSQ 96HS System as
described by the manufacturer (Biotage AB). Nucleotide
dispensation orders were chosen manually to generate
specific peaks for the targeted CYP2D6 and CYP2D8P
PCR products and to reveal any false amplification from
the CYP2D7P pseudogene. For the sequencing primer
A685FP, the dispensation order CGACACTC generated
CYP2D6-specific peaks at dispensations 5 (A) and 6 (C),
which were compared with the CYP2D8P-specific peaks
at dispensations 7 (T) and 8 (C) (see Fig. 2). For sequencing primer A1050FP, the dispensation order CTACATGCT generated CYP2D6-specific peaks at dispensations 5
(A) and 6 (T), which were compared with the CYP2D8Pspecific peak at dispensation 7 (G). The Sequence Analysis
software (Biotage AB) was used for measurement of peak
heights, and the results were exported to MS Excel for
further calculations and analysis. In the A685FP primer
assay, the CYP2D6-specific peak at dispensation 6 (C) and
the CYP2D8P-specific peak at dispensation 8 (C) were
used for the calculations. In the A1050FP primer assay, the
CYP2D6-specific peak at dispensation 6 (T) and the
CYP2D8P-specific peak at dispensation 7 (G) were used.
Results
assay design and optimization
Quantitative amplification assays were designed such
that the CYP2D6 gene copy number was examined with a
CYP2D6 pseudogene as reference. The choice of a reference gene with a high homology allowed one-tube amplification as well as sequencing of both genes with the same
PCR and sequencing primers. The pseudogene had two
important functions: (a) as a fixed and stable reference for
the quantification of the CYP2D6 copy number (having
the same amplification efficiency as the CYP2D6 target
gene); and (b) as a positive internal control for the assay
and the primers used for analysis of the CYP2D6 gene.
The pseudogene CYP2D7P was rejected as a reference
because it has been reported to be relatively frequently
duplicated (1 ). In contrast, neither deletions nor duplications have been reported for the CYP2D8P gene (which is
Clinical Chemistry 51, No. 3, 2005
525
Fig. 2. Typical Pyrograms from assays using the two sequencing primers A685FP and A1050FP on PCR products from samples differing in the
number of CYP2D6 genome copies.
The x axes are time axes showing the addition of nucleotides to the sequencing reaction. The y axes show light emission obtained as relative light units. nxD6 indicates
the number of CYP2D6 genome copies: (A), 0xD6; (B), 1xD6; (C), 2xD6; (D), 3xD6; (E), 4xD6. The left panels show Pyrograms from sequencing primer A685FP; the
right panels show Pyrograms from sequencing primer A1050FP. Arrows indicate peaks generated from nucleotide incorporations specific for CYP2D6 (thin arrows) and
CYP2D8P (thick arrows). These peaks were used in the gene copy number determination.
⬃95% homologous to CYP2D6 at the nucleotide level),
which qualified this gene as the reference.
The relative amounts of amplification products were
determined by use of the Pyrosequencing technology to
sequence through short regions that differed between the
CYP2D6 target gene and the reference gene. The signal
peak pattern, presented in a Pyrogram, was thus a product of two sequencing reactions that were performed
526
Söderbäck et al.: CYP2D6 Gene Copy Number
simultaneously. Three types of peaks occurred in the
Pyrogram: peaks specific for CYP2D6 (i.e., free from
interference by CYP2D8P); peaks specific for CYP2D8P
(i.e., free from interference by CYP2D6); and peaks consisting of a combined light signal from both genes. We
determined the number of CYP2D6 genes by calculating
the ratio between CYP2D6-specific peaks and peaks that
were specific for the CYP2D8P gene. All peaks in the
Pyrograms were, however, used to ensure the specificity
of the assay.
Several potential target regions in the CYP2D6 and
CYP2D8P genes were amplified by PCR and analyzed.
The majority of regions investigated were unstable and
caused various problems when tested on the DNA sample
panel, most probably because of variability within the
pseudogene sequence. One region obtained with PCR
primers A1058FP and A1051RPB at the end of exon 6,
however, was robust for the majority of DNA samples
tested (Fig. 1); the ratio was found to be proportional to
the number of CYP2D6 genes in the sample. Typical
results of genotyping using the sequencing primers
A685FP and A1050FP on this PCR product are shown in
Fig. 2. The designation “nxD6” indicates the number of
CYP2D6 genome copies. 0xD6 was a genotype that was
homozygous for the CYP2D6*5 allele, i.e., CYP2D6 was
deleted on both chromosomes and only the sequence from
CYP2D8P was seen. 1xD6 indicates a sample with one
CYP2D6 gene copy (heterozygous for CYP2D6*5), and the
expected peak ratio was ⬃0.5. 2xD6 indicates a wild-type
sample with two copies of CYP2D6, i.e., one on each
chromosome and a ratio of 1.
pcr annealing temperature
The effect of PCR annealing temperature on the specific
peak heights from CYP2D6 and CYP2D8P in a wild-type
genome (CYP2D6*1/*2) was tested in a gradient PCR with
annealing temperatures in the range 50 – 65 °C. The PCR
products were analyzed with the two sequencing primers,
and specific peak heights from the CYP2D6 and CYP2D8P
genes as well as the ratio between them were plotted as a
function of temperature (Fig. 3). Skewed peak heights
indicated differences in PCR efficiency between CYP2D6
and CYP2D8P at all PCR annealing temperatures except
for a relatively narrow window of 58 – 60 °C. At annealing
temperatures below this window, the CYP2D8P-specific
peak heights were higher than those for CYP2D6, whereas
the CYP2D6-specific peak heights were higher than those
for CYP2D8P at higher temperatures. The results were as
expected based on the PCR primer A1058FP sequence,
which has a C/T degeneracy between the CYP2D6 and
CYP2D8P sequences. An annealing temperature of 59 °C
was chosen to obtain a ratio close to 1.
pyrosequencing analysis
Peak heights exported from the Pyrosequencing software
were further analyzed in Microsoft Excel, where the
relative peak heights between CYP2D6- and CYP2D8P-
Fig. 3. Effect of PCR annealing temperature on peak heights.
(A and B), peak heights (in relative light units) specific for CYP2D6 (Œ) and
CYP2D8P (f) in a wild-type genome (CYP2D6*1/*2) as an effect of a gradient
annealing temperature from 50 to 65 °C. (A), sequencing primer A685FP; (B),
sequencing primer A1050FP. (C), ratios of specific peak heights (CYP2D6/
CYP2D8P) from the gradient PCR products. ⽧, primer A685FP; Œ, primer
A1050FP. The ratio is close to 1 at ⬃59 °C.
specific peaks were calculated. Panels A1 and A2 in Fig. 4
are scatter plots of the peak-height ratios for the respective sites for 40 selected samples of different genotypes,
showing conclusive centering of ratios for both positions
around values of 0 (0xD6), 0.5 (1xD6), 1 (2xD6), 1.5 (3xD6),
and higher ratios for three samples. According to our
analysis, two of these samples carried 4 gene copies, but
that has not been possible to verify by any other technique. The third sample carried even more than 4 gene
copies, indicated by the even higher peak-height ratio.
This is visualized more clearly if the ratios are sorted from
minimum to maximum (see panels B1 and B2 in Fig. 4).
Note that samples exhibiting any of the variants described
in the section below were excluded from these plots. The
peak-height ratios of these samples were not relevant, and
Clinical Chemistry 51, No. 3, 2005
527
.
.
.
.
.
.
.
.
.
.
.
.
Fig. 4. Scatter plots of CYP2D6-/CYP2D8-specific peak-height ratios.
The y axes show the calculated ratio between CYP2D6- and CYP2D8P-specific peaks. The x axes are the serial numbers of the samples. (A1 and B1), ratios for
sequencing primer A685FP. (A2 and B2), ratios for sequencing primer A1050FP (same samples as in panels A1 and B1). Panels A1 and A2 show the DNA samples
in order of collection; panels B1 and B2 show the samples when sorted from lowest to highest ratios based on ratios from the A1050FP assay. Observe that the
frequencies of the different genotypes are from a nonrepresentative subset of samples.
the two peak ratio values from such samples would
segregate drastically, indicating the variation and the
requirement for visual examination of the Pyrograms.
reproducibility
To study the reproducibility of the assays, we repeatedly
analyzed four samples with known CYP2D6 gene copy
numbers (1xD6, 2xD6, 3xD6, and 4xD6) on nine separate
occasions (on different days) and plotted the D6/D8P
ratio against the expected number of CYP2D6 copies (Fig.
5). Data points from samples analyzed on a specific
occasion (day) are connected by a line, and the linear
regression coefficients (R2) were calculated for each such
series. For the A685FP assay, the linear regression coefficients varied from 0.9731 to 0.9994, and for A1050FP, they
varied between 0.9632 and 0.9979. The lines confirmed a
good linearity in the analysis of copy numbers, and the
comparison between lines indicated high reproducibility.
This analysis also indicated that the assay using sequencing primer A685FP gave a somewhat greater day-to-day
variation than the assay using the A1050FP primer.
samples generating atypical pyrosequencing data
The CYP2D8P gene deviated from the expected sequence
in 13 of the ⬃200 test samples. One relatively common
allelic type with a C3 T transition at the equivalent of
528
Söderbäck et al.: CYP2D6 Gene Copy Number
(compare panel B1 in Fig. 6 with panel B1 in Fig. 2). This
sample exhibited a typical 1xD6 pattern when analyzed
with the A1050FP sequencing primer (compare panel B2
in Fig. 6 with panel B2 in Fig. 2).
Another allelic variant was found in one 3xD6 sample
when analysis was performed with the A685FP sequencing primer. This variant was more difficult to interpret.
The Pyroseqencing data (Pyrogram) from the A685FP
assay exhibited a pattern indicating multiple copies
(6xD6) (panel C1 in Fig. 6). In contrast, the Pyrogram from
the A1050FP assay exhibited a typical 3xD6 pattern (panel
C2 in Fig. 6). The only plausible explanation is that the
sequencing primer A685FP did not anneal to one of the
CYP2D8P alleles in this sample, leading to CYP2D8Pspecific peaks being generated from only one of the two
CYP2D8P alleles, giving reference peaks one half the
height of those obtained from the same sample analyzed
with the A1050FP primer.
assay verification
Fig. 5. Analysis of reproducibility of the CYP2D6 gene copy number
assays.
(A), A685FP assay; (B), A1050FP assay. Data points for samples analyzed on a
specific occasion (day) are connected by a line.
position 2933 in CYP2D6 was found downstream of the
A1050FP sequencing primer. This was readily revealed,
however, by the unexpected Pyrogram pattern (compare
panel A2 in Fig. 6 with panel C2 in Fig. 2), which was
affected by the signals from the variants of the CYP2D8P
gene. All of these samples exhibited a typical 2xD6
pattern when analyzed with the A685FP sequencing
primer (compare panel A1 in Fig. 6 with panel C1 in Fig.
2). One 1xD6 sample with a base transition in one of the
alleles, rendering an equivalent effect for the A685FP
primer, was also found among the 200 test samples
The two sequencing primers A685FP and A1050FP gave
interpretable results in 269 of the 270 pregenotyped DNA
samples. A few samples had to be reanalyzed because of
discrepancies between the A685FP and A1050FP assays.
Thirteen samples (4.8%) were identified as carrying one
allele with a duplication of CYP2D6 as indicated by the
gene copy number, 3xD6 (Table 1). Long-range PCR
analysis verified these findings.
The gene copy number 0xD6 was not found in any of
the samples from blood donors. Pyrograms corresponding to a gene copy number of 1xD6 were generated by
23 of 24 samples from individuals previously genotyped
as heterozygous carriers of the CYP2D6*5 allele. The
aberrant sample had a gene copy number of 2xD6, which
did not compare with the previously found genotype,
CYP2D6*4/*5. The sample was reanalyzed with the same
result.
A typical allelic distribution of the CYP2D6 gene
(2xD6) was found in 231 samples. In the first set of
analyses, three samples generated results that deviated
between the two primers. The A685FP primer indicated
3xD6, whereas the A1050FP primer indicated 2xD6 in two
of these samples. The third sample generated the opposite
aberrant result. After reanalysis, all three samples gave
the result 2xD6. This was also verified by the long-range
PCR for multiple genes. We could find no explanation for
the aberrant results in the first analysis.
One sample pregenotyped as wild-type, CYP2D6*2/*2,
was identified as 1xD6 by the Pyrosequencing assay.
However, reinvestigation using long-range PCR did not
verify the 1xD6 gene copy number, i.e., no gene deletion
was observed. Of the samples from 269 blood donors, four
2xD6 samples and one 3xD6 sample were found to carry
one allele with the variant CYP2D8P sequence identified
with primer A1050FP (see the Pyrogram in Fig. 6). A
typical peak pattern was obtained with the sequencing
primer A685FP.
529
Clinical Chemistry 51, No. 3, 2005
Fig. 6. Pyrograms from samples generating a typical pattern in one of the assays but not in the other.
The x axes show the addition of nucleotides, and the y axes show the light signal obtained as relative light units. The left panels show the A685FP primer assay; the
right panels show the A1050FP primer assay. Sample A (top row) shows a typical 2xD6 pattern in the A685FP assay, but an atypical 2xD6 pattern in the A1050FP assay.
Sample B (middle row) shows a typical 1xD6 pattern in the A1050 assay and an atypical pattern in the A685 assay. Sample C (bottom row) shows a typical 3xD6 pattern
in the A1050 assay and a pattern resembling 6xD6 in the A685 assay. Arrows indicate peaks generated from nucleotide incorporations specific for CYP2D6 (thin
arrows) and CYP2D8P (thick arrows).
Discussion
Table 1. Numbers of CYP2D6 gene copies found in 269 of
270 blood donor samples tested.
Gene copy numbera
Genotype
*1/*1
*1/*2
*1/*3
*1/*4
*1/*5
*1/*6
*2/*2
*2/*3
*2/*4
*2/*5
*3/*4
*4/*4
*4/*5
*4/*6
*5/*6
Total
a
0xD6
1xD6
2xD6
29
68
2
47
3xD6
3
2
6
9
1
2
24
3
41
1
1
7
6
0
1
24
2
11
1
2
232
13
Two samples with disparate results between previous genotyping and
Pyrosequencing analysis are indicated in bold.
In this study, we describe an assay for determination of
the CYP alleles CYP2D6*5 and CYP2D6*2xN. The assay,
which is based on competitive PCR between the CYP2D6
and CYP2D8P genes and Pyrosequencing analysis, generated stable patterns and unequivocal genotype determination for samples carrying the CYP2D6*5 allele in both a
homozygous and heterozygous state. For samples carrying a multiple-gene-copy allele (CYP2D6*2xN), it was
possible to identify carriers of 3 or 4 gene copies. The
samples carrying 4 gene copies were not verified with
other techniques. However, this conclusion was supported by the high linearity in the reproducibility assay
(Fig. 5). The resolution of alleles carrying multiple gene
copies may be increased by further optimization of the
method. Jansson et al. (35 ) have also successfully applied
Pyrosequencing technology for gene copy analysis of the
glutathione S-transferase M1 gene and identified an allele
(GSTM1*0) with complete deletion of the gene. A related
approach has been described by Pielberg et al. (36 ), who
used breakpoint analysis, which cannot be applied on
CYP2D6 because the breakpoints are located in the 2.8-kb
CYP-REP region. The analysis of two positions on the
530
Söderbäck et al.: CYP2D6 Gene Copy Number
same PCR product, as described in the current method,
offers substantial advantages because it facilitates the
resolution of ambiguous results.
DNA samples prepared from blood in four different
laboratories were analyzed in this study. Interestingly,
small differences in the relative peak patterns of the
sequence were observed from DNA samples delivered
from the different laboratories and purified by different
methods (data not shown). Our assumption is that the
method for preparation of genomic DNA may influence
the first cycles of the PCR reaction, perhaps via remaining
material (e.g., histone complexes or regulatory proteins)
bound to the DNA after the preparation. This difference
in relative peak pattern was more pronounced when we
used primer batches from a certain vendor (data not
shown). The conclusion is that care must be taken when
interpreting variability that may be attributable to low
quality of the primers or the DNA preparation used.
A few allelic variations found within the CYP2D8P
gene generated atypical Pyrogram patterns. A simple
solution for these problems would be to use a patternrecognition approach for the genotyping analysis. Because stable patterns were generated for the different
genotypes as well as for the atypical samples, this assay is
suitable for software-based recognition of Pyrosequencing peak patterns. This would facilitate the analysis step
considerably because export of peak heights and Excelbased analysis would be unnecessary. A relatively simple
prototype software for such purposes was tested with
very good results.
Blind tests of DNA from blood donors showed rare
and unexplained discrepancies in results generated with
the different techniques used. Reanalysis of the samples
clarified most of these discrepancies. However, two samples remained divergent. One sample previously typed as
CYP2D6*4/*5 gave a gene copy number of 2xD6, which is
not usually found for the actual genotype. A possible
explanation is that the *4 allele in this case contained a
gene duplication. This is also supported by the long-PCR
analysis for multiple genes, which showed a duplication
of one of the alleles. In the other sample, the genotype
CYP2D6*2/*2 predicted a wild-type 2xD6 gene copy number, but instead we found a 1xD6 profile. This was the
only sample of the 269 blood samples tested for which a
discrepancy remained between the found copy number
and the genotype in combination with long-PCR methods
for the CYP2D6*5 allele and multiple CYP2D6 genes. One
sample from the blood donors was excluded because of
recurrently unclear results from both Pyrosequencing and
long-PCR reactions. Considering the relatively high probability for variation in the CYP2D8P pseudogene, it could
be assumed that other allelic variants will appear in larger
sample groups. However, the variations identified here
all indicate the strength of the assay, which generates
sequence background from two positions. It is possible to
determine the reason for unexpected (atypical) patterns in
one position by having as a key the typical appearance of
peaks generated at the other position.
A possible problem with the assay would be total or
partial deletion of one of the two CYP2D8P alleles in a
particular genome. To our knowledge, a sample carrying
such a deletion has not been reported. Such a sample
would generate Pyrograms for both positions with
CYP2D8P-specific reference peaks that are one half of the
expected height relative the CYP2D6-specific peaks. The
gene quantification analysis would suggest an erroneous
4xD6 genotype for wild-type (2xD6) samples. The same
type of problem would appear if the variation within
CYP2D8P was so pronounced that the PCR reaction does
not work for one of the CYP2D8P alleles. As described
above, such a nonfunctional primer sequence most probably contributed to the aberrant result obtained during
sequencing of one of the tested samples. If both CYP2D8P
alleles are deleted in a sample, the specific CYP2D8P
peaks will be absent from the Pyrogram, indicating the
necessity to use other techniques (i.e., long-range PCR) for
the analysis of CYP2D6 gene copy number. No such
samples were found among those tested.
Except for long-range PCR, all other methods for the
determination of CYP2D6 gene copy number that have
been reported in the literature use an unrelated gene as
reference. To our knowledge, this is the first report
describing an analysis using a highly related gene as
reference. There are several advantages associated with
the use of a related gene and coamplification in a method
such as the one described here. Only a single set of
primers is used (the same for both target and reference
gene), which increases the probability of an accurate
determination of gene copy number. The generation of
amplicons from unrelated genes may lead to variations in
amplification efficiencies between the different genetic
regions. The use of a single set of primers also enables
more straightforward method validation. Both of these
advantages should make the approach described here
attractive for pharmacogenetic studies.
We are grateful for the receipt of genotyped DNA samples from Drs. Steven Wong and Paul Jannetto (Department of Pathology, Medical College of Wisconsin, Milwaukee, WI), Dr. Inger Johansson (Division of Molecular
Toxicology, Karolinska Institute, Stockholm, Sweden),
and Dr. Mia Wadelius (Department of Clinical Pharmacology, Akademic Hospital, Uppsala, Sweden). We are
also grateful to Ann-Britt Gladh for her excellent work.
This work was supported by Grant 03-15 from The
National Board of Forensic Medicine.
References
1. Online Mendelian Inheritance in Man (OMIM) http://www.ncbi.
nlm.nih.gov/entrez/Omim (accessed May 24, 2004).
2. Kimura S, Umeno M, Skoda RC, Meyer UA, Gonzalez FJ. The
human debrisoquine 4-hydroxylase (CYP2D) locus: sequence and
Clinical Chemistry 51, No. 3, 2005
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
identification of the polymorphic CYP2D6 gene, a related gene,
and a pseudogene. Am J Hum Genet 1989;45:889 –904.
Bertilsson L, Dahl M-L, Dalen P, Al-Shurbaji A. Molecular genetics
of CYP2D6: clinical relevance with focus on psychotropic drugs.
Br J Clin Pharmacol 2002;53:111–22.
Home Page of the Human Cytochrome P450 (CYP) Allele Nomenclature Committee. CYP2D6 allele nomenclature. http://www.
imm.ki.se/CYPalleles/cyp2d6.htm (accessed May 24, 2004).
Heim MH, Meyer UA. Evolution of a highly polymorphic human
cytochrome P450 gene cluster: CYP2D6. Genomics 1992;14:
49 –58
Johansson I, Lundqvist E, Bertilsson L, Dahl ML, Sjoqvist F,
Ingelman-Sundberg M. Inherited amplification of an active gene in
the cytochrome P450 CYP2D locus as a cause of ultrarapid
metabolism of debrisoquine. Proc Natl Acad Sci U S A 1993;90:
11825–9.
Hersberger M, Marti-Jaun J, Rentsch K, Hänseler E. Rapid detection of the CYP2D6*3, CYP2D6*4, and CYP2D6*6 alleles by
tetra-primer PCR and of the CYP2D6*5 allele by multiplex long
PCR. Clin Chem 2000;46:1072–7.
Dalen P, Dahl ML, Ruiz ML, Nordin J, Bertilsson L. 10-Hydroxylation of nortriptyline in white persons with 0, 1, 2, 3, and 13
functional CYP2D6 genes. Clin Pharmacol Ther 1998;63:444 –52.
Dahl ML, Johansson I, Bertilsson L, Ingelman-Sundberg M, Sjoqvist F. Ultrarapid hydroxylation of debrisoquine in a Swedish
population. Analysis of the molecular genetic basis. J Pharmacol
Exp Ther 1995;274:516 –20.
Scordo MG, Spina E, Facciola G, Avenoso A, Johansson I, Dahl
ML. Cytochrome P450 2D6 genotype and steady state plasma
levels of risperidone and 9-hydroxyrisperidone. Psychopharmacology (Berl) 1999;147:300 –5.
Aklillu E, Persson I, Bertilsson L, Johansson I, Rodrigues F,
Ingelman-Sundberg M. Frequent distribution of ultrarapid metabolizers of debrisoquine in an Ethiopian population carrying duplicated and multiduplicated functional CYP2D6 alleles. J Pharmacol
Exp Ther 1996;278:441– 6.
Johansson I, Lundqvist E, Dahl M-L, Ingelman-Sundberg M. PCRbased genotyping for duplicated and deleted CYP2D6 genes.
Pharmacogenetics 1996;6:351–5.
Neville M, Selzer R, Aizenstein B, Maguire M, Hogan K, Walton R,
et al. Characterization of cytochrome P450 2D6 alleles using the
Invader system. Biotechniques 2002;32:S34 –S43.
Schaeffeler E, Schwab M, Eichelbaum M, Zanger UM. CYP2D6
genotyping strategy based on gene copy number determination by
TaqMan real-time PCR. Hum Mutat 2003;22:476 – 85.
Ronaghi M, Uhlen M, Nyren P. A sequencing method based on
real-time pyrophosphate. Science 1998;281:363, 365.
Berg LM, Sanders R, Alderborn A. Pyrosequencing technology and
the need for versatile solutions in molecular clinical research
[Review]. Expert Rev Mol Diagn 2002;2:361–9.
Ahmadian A, Gharizadeh B, Gustafsson AC, Sterky F, Nyren P,
Uhlen M, et al. Single-nucleotide polymorphism analysis by pyrosequencing. Anal Biochem 2000;280:103–10.
Alderborn A, Kristofferson A, Hammerling U. Determination of
single-nucleotide polymorphisms by real-time pyrophosphate DNA
sequencing. Genome Res 2000;10:1249 –58.
Garcia CA, Ahmadian A, Gharizadeh B, Lundeberg J, Ronaghi M,
Nyren P. Mutation detection by pyrosequencing: sequencing of
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
531
exons 5– 8 of the p53 tumor suppressor gene. Gene 2000;253:
249 –57.
Goriely A, McVean GA, Rojmyr M, Ingemarsson B, Wilkie AO.
Evidence for selective advantage of pathogenic FGFR2 mutations
in the male germ line. Science 2003;301:643– 6.
Jonasson J, Olofsson M, Monstein HJ. Classification, identification and subtyping of bacteria based on Pyrosequencing and
signature matching of 16S rDNA fragments. APMIS 2002;110:
263–72.
Rahim A, Coutelle C, Harbottle R. High-throughput Pyrosequencing
of a phage display library for the identification of enriched
target-specific peptides. Biotechniques 2003;735:317–24.
Gruber JD, Colligan PB, Wolford JK. Estimation of single nucleotide polymorphism allele frequency in DNA pools using Pyrosequencing. Hum Genet 2002;110:395– 401.
Neve B, Froguel P, Corset L, Vaillant E, Vatin V, Boutin P. Rapid
SNP allele frequency determination in genomic DNA pools by
Pyrosequencing. Biotechniques 2002;32:1138 – 42.
Lavebratt C, Sengul S, Jansson M, Schalling M. Pyrosequencingbased SNP allele frequency estimation in DNA pools. Hum Mutat
2004;23:92–7.
Tost J, Dunker J, Gut IG. Analysis and quantification of multiple
methylation variable positions in CpG islands by Pyrosequencing.
Biotechniques 2003;35:152– 6.
Colella S, Shen L, Baggerly KA, Issa JP, Krahe R. Sensitive and
quantitative universal Pyrosequencing methylation analysis of
CpG sites. Biotechniques 2003;35:146 –50.
Eriksson S, Berg L, Wadelius M, Alderborn A. Cytochrome P450
genotyping by multiplexed real-time DNA sequencing with Pyrosequencing technology. Assay Drug Dev Technol 2002;1:49 –59.
Zackrisson AL, Lindblom B. Identification of CYP2D6 alleles by
single nucleotide polymorphism analysis using pyrosequencing.
Eur J Clin Pharmacol 2003;59:521– 6.
Wadelius M, Sorlin K, Wallerman O, Karlsson J, Yue QY, Magnusson PK, et al. Warfarin sensitivity related to CYP2C9, CYP3A5,
ABCB1 (MDR1) and other factors. Pharmacogenomics J 2004;4:
40 – 8.
Furman KD, Grimm DR, Mueller T, Holley-Shanks RR, Bertz RJ,
Williams LA, et al. Impact of CYP2D6 intermediate metabolizer
alleles on single-dose desipramine pharmacokinetics. Pharmacogenetics 2004;14:279 – 84.
Min DI, Ellingrod VL, Marsh S, McLeod H. CYP3A5 polymorphism
and the ethnic differences in cyclosporine pharmacokinetics in
healthy subjects. Ther Drug Monit 2004;26:524 – 8.
Gustincich S, Manfioletti G, Del Sal G, Schneider C, Carninci P. A
fast method for high-quality genomic DNA extraction from whole
human blood. Biotechniques 1991;11:298 –300, 302.
Lovlie R, Daly AK, Molven A, Idle JR, Steen VM. Ultrarapid
metabolizers of debrisoquine: characterization and PCR-based
detection of alleles with duplication of the CYP2D6 gene. FEBS
1996;392:30 – 4.
Jansson M, Rada A, Tomic L, Larsson L-I, Wadelius C. Analysis of
the Glutathione S-transferase M1 gene using pyrosequencing and
multiplex PCR—no evidence of association to glaucoma. Exp Eye
Res 2003;77:239 – 43.
Pielberg G, Day AE, Plastow GS, Andersson L. A sensitive method
for detecting variation in copy numbers of duplicated genes.
Genome Res 2003;13:2171–7.